Roger Mills <[log in to unmask]> writes:

> There are two classes ("genders") of nouns:  1) animate:  all living things
> capable of motion, including the Sun, Moons, the spirits,  'planet' and
> certain violent weather conditions (but excluding microscopic things like
> cells, bacteria)   2) neuter:  everything else.
> These are declined with suffixes through 4 cases-- Nom. Gen. Dat. Acc.-- and
> 2 numbers sing. and plur.  The suffixes are:
> Singular:
>   Nom. - nothing.  bases may end in any cons. or vowel
>   Gen. -  -i   (The writing system required a glide between ...V-i)
>   Dat. -  -e   (ditto)
>   Acc. -  anim. -(V)n*     neut:  nothing; neut. Nom. and Acc. are
>                                   identical in both sing. and plur.
> Plural (anim. and neut. differ):
>      Anim.    Nom.   -(i)la           Neut:  -(V)sh*
>               Gen.   -(i)li                  -(V)shi
>               Dat.   -(i)le                  -(V)she
>               Acc.   -(i)lan                 -(V)sh

[table reformatted in order to get things in line]

I notice that (similar to the way it is in Indo-European), neuters, i.e.
inanimate nouns, have morphologically identical nominatives and
accusatives.  Is this a trace of a former active system (as some
linguists assume for Proto-Indo-European)?

By "active system" I mean the following:

Intransitive verbs fall into two classes - active and non-active.
Active verbs denote actions performed by the subject, such as "to walk"
or "to laugh"; non-active verbs denote either states, e.g. "to sit", or
events happening to the subject, e.g. "to fall".  Of these two classes,
active verbs treat their subject like a transitive subject, while
non-active verbs treat it like an object.


The child throws the stone.
    AGT              OBJ

The child laughs.

The stone falls.

The abbreviations refer to the terms "agentive" and "objective", which
are the names I assigned to the cases in my active-type conlang,

Inanimate nouns, however, rarely ever are agents (and if they are,
closer inspection usually reveals them as pseudo-agents, as in the
example "The stone smashes the window", where it is actually such that
some external force set the stone in motion and caused it to smash the
window.  Hence, in some language they never occur in agentive case -
that case is simply missing from the inanimate paradigm.  This is so in
Nur-ellen, for example.

My question is whether in the conhistory of Kash, there had been such a
system in the past, which later gave way to the more familiar nom/acc
system by using the former agentive as a nominative, i.e. also for
non-active intransitive subjects, and restricting the former objective
to the role of an accusative.  When this happened, a new nominative had
to be invented for inanimate nouns because those lacked an agentive to
draw upon, and the objective was used for both nominative and
accusative, which are thus morphologically identical.  This kind of
development has been proposed for Proto-Indo-European (by someone named
Winfried P. Lehmann); I wonder whether it happened that way in Kash?